Acid mine drainage (AMD) is a global problem that may have serious human health and environmental implications. Laboratory and field tests are commonly used for predicting AMD, however, this is challenging since its formation varies from site-to-site for a number of reasons. Furthermore, these tests are often conducted at small-scale over a short period of time. Subsequently, extrapolation of these results into large-scale setting of mine sites introduce huge uncertainties for decision-makers. This study presents machine learning techniques to develop models to predict AMD quality using historical monitoring data of a mine site. The machine learning techniques explored in this study include artificial neural networks (ANN), support vector machine with polynomial (SVM-Poly) and radial base function (SVM-RBF) kernels, model tree (M5P), and K-nearest neighbors (K-NN). Input variables (physico-chemical parameters) that influence drainage dynamics are identified and used to develop models to predict copper concentrations. For these selected techniques, the predictive accuracy and uncertainty were evaluated based on different statistical measures. The results showed that SVM-Poly performed best, followed by the SVM-RBF, ANN, M5P, and KNN techniques. Overall, this study demonstrates that the machine learning techniques are promising tools for predicting AMD quality.
Investigations of the geochemistry of inactive pyritic uranium tailings in the Elliot Lake Mining district of Ontario have focused on the Nordic tailings management area, where two impoundments are located in natural bedrock basins. The tailings are 8-12 m thick and overlie a localized deposit of glaciofluvial sands. Analyses of the solid, liquid, and gas phases in the vadose zone of the tailings show that gas-phase oxygen levels drop rapidly within 0.7 to 1.5 m of the surface, indicating rapid oxygen consumption during pyrite oxidation. Oxidation during the past 15 to 20 years has caused a marked depletion of near-surface pyrite. The oxidation of pyrite in the vadose zone imparts to infiltrating precipitation high concentrations of Fe, SO42-, various heavy metals, and a pH generally between 1.5 and 4. The acidic infiltration moves downward at a rate of 0.2 to 2.0 m/yr, displacing high-pH groundwater that originated as process water discharged from the mill. It now occupies the entire tailings thickness over a small area of the tailings. At one location a well-defined plume of high-Fe2+ tailings-derived groundwater has developed in the sand aquifer adjacent to the tailings. The plume consists of three zones: an inner core characterized by Fe > 5000 mg/L, pH < 4.8, and elevated concentrations of several heavy metals and radionuclides; an outer zone with Fe < 2500 mg/L, pH > 5.5, and relatively low concentrations of heavy metals and radionuclides; and a transition zone separating the two. Although the average linear groundwater velocity is about 440 m/yr near the dam, reactions such as mineral dissolution, precipitation and coprecipitation retard the migration of the front of the inner core, producing an observed frontal migration rate of approximately 1 m/yr. Groundwater from the outer zone of the plume flows laterally towards a small stream, where a portion of it is now discharging into the stream bed. The discharge results in the precipitation of amorphous ferric hydroxide on the stream bed. Most of the H+ produced by Fe precipitation is buffered, and only a moderate decrease in stream pH is observed. Inner zone conditions will not reach the stream unless input of low-pH groundwater from the tailings continues for several hundred years. Although the rate of pyrite oxidation in the Nordic Main tailings has been decreasing, there is sufficient pyrite in the tailings to generate high-Fe groundwater for several decades or more. Calculated groundwater migration rates indicate that in the next few decades acidic, low-pH groundwater will occupy the entire tailings thickness over most of the tailings area, causing an increase in the total flux of contaminated groundwater into the underlying aquifer. The outer zone of the plume has already arrived at a small stream, and acidification of the surface waters may increase if the Fe concentration in the groundwater seepage increases.
Mine-rock piles are complex hydrogeologic systems. As a result, the current knowledge of their physical and chemical hydrogeology is too limited to permit accurate predictions of water chemistry through time based on detailed simulations of their internal processes. However, a simplistic empirical model based on general knowledge and available data can be used to obtain rough estimates of seepage chemistry through time. This empirical model is based on five factors: (1) the production rates of metals, nonmetals, acidity, and alkalinity under acid and pHneutral conditions, (2) the volume rate of flow through the rock pile based on infiltration of precipitation, (3) the elapsed time between infiltration events, (4) the residence time of the water within the rock pile, and (5) the percentage of mine rock in the pile flushed by the flowing water. The last factor is most difficult to define at many minesites, but can be assumed due to its apparently frequent narrow range of 5 to 20%. A hypothetical example illustrates how to use the model and highlights other complications like secondary-mineral precipitation that may also have to be considered. A field example based on data from an actual minesite demonstrates the accuracy of the model as compared to measured concentrations. Again, this model ignores many complexities of mine-rock piles and is thus only useful for rough estimates of future chemistry.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.